Chronic Stress Burden and Genes Together Shape Lung Cancer Risk, Study Finds

For decades, scientists have suspected that the slow grind of chronic stress leaves fingerprints on the body that go far beyond a racing heart or a sleepless night. A new study published in PLOS Digital Health offers some of the most concrete evidence yet that those fingerprints may extend all the way to cancer. By following more than 336,000 adults in the United Kingdom for years, researchers found that people carrying a heavy burden of what biologists call allostatic load — a composite measure of wear and tear across multiple physiological systems — were significantly more likely to develop lung cancer than their peers with normal readings. The effect was not modest. Those in the highest category of allostatic load faced roughly a 63 percent higher risk of lung cancer compared with the normal group, and the association was strongest for small-cell lung cancer, the most aggressive and least common of the major tumor subtypes.

Allostatic load is a concept borrowed from stress physiology, first articulated in the 1990s to describe the cumulative physiological cost of repeated adaptation to adversity. Rather than measuring stress as a feeling, it aggregates objective markers drawn from routine clinical data: blood pressure, body mass index, waist-to-hip ratio, cholesterol fractions, blood glucose, inflammatory markers such as C-reactive protein, and indicators of organ function across the cardiovascular, metabolic, immune, and endocrine systems. When many of these markers drift into abnormal territory at the same time, the body is effectively operating in a state of sustained emergency. The researchers behind the new study treated this composite score as a single, quantifiable index of physiological dysregulation, and asked a deceptively simple question: does it predict lung cancer independently of smoking, age, and the other usual suspects?

The answer, drawn from a prospective analysis of 336,945 UK Biobank participants, appears to be yes. Each one-unit increase in the allostatic load score was associated with a 7.3 percent increase in lung cancer risk, with a hazard ratio of 1.073 and a 95 percent confidence interval of 1.053 to 1.094. When participants were grouped into categories, those classified as low allostatic load had a 28.4 percent higher risk than the normal group, while the high group carried a 62.9 percent higher risk, with a hazard ratio of 1.629. Crucially, the team examined histological subtypes separately and found the signal was not uniform. Small-cell lung cancer showed the strongest association, while lung adenocarcinoma and squamous cell carcinoma also displayed significant links. That pattern is intriguing because small-cell tumors are tightly tied to neuroendocrine biology and heavy tobacco exposure, raising questions about how systemic physiological stress might intersect with those pathways.

Genetics, of course, is the other half of the risk equation. The investigators constructed a polygenic risk score — a weighted sum of many common genetic variants each contributing a small increment of susceptibility — to capture each participant’s inherited predisposition to lung cancer. They then asked whether allostatic load and genetic risk acted independently, or whether one amplified the other. The result was striking in its clarity: the two factors combined multiplicatively in the data but showed no evidence of statistical interaction. In other words, they stack rather than synergize. Participants who sat in the high-allostatic-load group and simultaneously carried high genetic risk faced a 136.9 percent increase in lung cancer risk — a hazard ratio of 2.369, with a confidence interval of 1.565 to 3.583 — roughly two and a half times the risk of comparable individuals without either burden.

That finding has immediate practical implications for screening. Low-dose computed tomography, or LDCT, is the only screening modality proven to reduce lung cancer mortality, but eligibility criteria — typically built around age, smoking intensity, and smoking duration — miss a meaningful fraction of people who ultimately develop the disease. The researchers tested whether adding allostatic load to the existing eligibility framework improved risk discrimination, and it did. The composite stress measure provided incremental stratification beyond established LDCT screening criteria, suggesting that a panel of cheap, routinely collected clinical markers could help flag high-risk individuals whom current guidelines overlook. Because allostatic load is calculated from data already gathered in ordinary checkups, the barrier to testing this idea in clinical settings is far lower than for most novel biomarkers.

But the study did not stop at epidemiology. One of its most ambitious components was an attempt to peer inside the black box connecting cumulative physiological dysregulation to carcinogenesis, using plasma proteomics — the large-scale measurement of circulating proteins. Integrating proteomic data with machine learning, pathway enrichment, and mediation analyses, the team identified 111 proteins whose circulating levels were associated with allostatic load. Within that set, 34 proteins showed potential mediation signal, meaning their statistical behavior was consistent with the idea that they might lie on the causal path between physiological dysregulation and lung cancer. Among the leading candidates were hepatocyte growth factor (HGF), chemerin (RARRES2), and growth differentiation factor 15 (GDF15) — molecules with well-documented roles in inflammation, metabolism, cell growth, and stress signaling.

The identity of these candidate mediators is scientifically telling. HGF is a potent growth factor involved in tissue repair and cell migration, and it has been implicated in tumor invasion and metastatic behavior across multiple cancers. RARRES2, also known as chemerin, is an adipokine — a signaling molecule released by fat tissue — that links metabolic state to immune function, fitting neatly with the idea that metabolic dysregulation is a central pillar of allostatic load. GDF15 is one of the most consistently elevated proteins in contexts of tissue stress, inflammation, and aging, and elevated circulating GDF15 has been repeatedly associated with poor health outcomes and cancer presence. That all three emerged from an unbiased, data-driven screen lends biological plausibility to the epidemiological association, pointing toward inflammatory, metabolic, and growth-factor-related processes as the connective tissue between chronic stress physiology and lung carcinogenesis.

The authors are careful, appropriately, to frame the proteomic findings as exploratory. Mediation analysis in observational data can suggest pathways but cannot prove causation, and reverse causation — early, undetected tumors altering circulating protein levels — remains a consideration in any prospective cancer study. Still, the design has notable strengths. The cohort was large and prospective, meaning allostatic load was measured before cancer diagnosis, which reduces the likelihood that the disease drove the biomarker rather than the reverse. The study also adjusted for established confounders and examined subtype-specific outcomes, which is where some of the most interesting biology surfaced. The pronounced association with small-cell lung cancer, in particular, invites follow-up work on how chronic inflammatory and endocrine stress might influence neuroendocrine differentiation in the lung.

What makes the study resonate beyond lung cancer specialists is its conceptual reframing of risk. Lung cancer risk assessment has long been dominated by behavioral exposure — chiefly smoking — and inherited genetics. Allostatic load sits at the intersection of those factors and everything else that shapes physiological wear: occupational strain, socioeconomic adversity, sleep, diet, chronic disease, and the accumulated micro-stresses of daily life. A composite biomarker of this kind does not identify a single culprit; instead, it captures the integrated output of many upstream causes. That is both its strength and its limitation. It offers a powerful summary signal for risk stratification, but it does not by itself tell clinicians which intervention — stress reduction, metabolic management, blood pressure control — would most effectively lower a given patient’s risk.

The road from association to clinical practice runs through replication and, ultimately, intervention trials. Future studies will need to confirm these findings in independent cohorts with different demographic profiles, refine how allostatic load is scored, and test whether incorporating it into screening algorithms genuinely reduces lung cancer mortality rather than merely reshuffling risk categories. The proteomic mediators identified here — HGF, RARRES2, GDF15, and their 31 companions — now constitute a testable biological hypothesis, one that laboratory models and interventional studies can interrogate directly. If the findings hold, the implications are considerable: a measurable, modifiable, and routinely computable indicator of cumulative physiological burden could join age and smoking history in the calculus of who deserves a closer look for the world’s leading cause of cancer death. For now, the study stands as a compelling demonstration that the body’s long memory of stress, written in blood pressure, metabolism, and inflammation, may be legible in cancer risk as well.

Subject of Research: The association of allostatic load and genetic susceptibility with lung cancer risk and its proteomic mediators

Article Title: Allostatic load, genetic susceptibility, and lung cancer risk: Evidence from a prospective cohort and integrative proteomic analyses

Article References: Yang, F., Wang, W., Chen, F., Wang, Z., Wang, Y., Qi, T., He, Z., Liang, S., Bu, Q., Qiu, X., Jiaqing, D., Ren, S., Dai, J., & Wang, X. (2026). Allostatic load, genetic susceptibility, and lung cancer risk: Evidence from a prospective cohort and integrative proteomic analyses. PLOS Digital Health, 5(10), e0001766. https://doi.org/10.1371/journal.pdig.0001766

Image Credits: AI Generated

DOI: 10.1371/journal.pdig.0001766

Keywords: allostatic load, lung cancer, UK Biobank, polygenic risk score, proteomics, small-cell lung cancer, LDCT screening, chronic stress, GDF15, HGF, risk stratification, PLOS Digital Health